Been wanting to do a little write up for a few days about a point I made some time ago; I may suck a bit at explaining myself, but here we go. I'll probably do it in 2 posts, 1 for each state.
We're going to analyse just 2 states, Florida and NC. Why these? Because a) they matter b) they have good pollster data we can grab.
We'll start with NC.
The 7 largest counties have between a 44-52% average turnout so far. Of total registered voters. All voted blue in 2016. 1.3 million voters (this time). Around 1.85 total in 2016. We'll call this set 1
The next 6 largest counties have a minimum turnout of around 43%, a maximum of 50%. Again of total registered voters. All voted red in 2016. 400k voters this time around. 570k total voters in 2016. We'll call this set 2
We'll add some granularity soon, but those are the headlines.
Let's take a look at some good polling: The 3 I have selected are Siena/NYT, Monmouth, and ABC/Wash Post
First we shall challenge our pollsters on turnout:
A common question asked (may vary in wording, this wording stolen from Siena/NYT) Regardless of whether you intend to vote in the November general election, would you vote in person on Election Day, vote in person before Election Day, or vote by mail?
We will amalgamate 'vote by mail' and 'early voting' for ease of reading: [may not add up to 100% as I'll disregard undecideds. Note that the majority in NC don't vote by mail, but early.
Siena/NYT: On the day, 40%. Early: 57%
Monmouth: On the day: 33%. Early: 64%
ABC/Wash Post: On the day: 29%. Early 66%
Certainly Trumps aggressive campaign against mail voting has worked. 50% of voters in NC will vote in person, but early.
Let's now plug our numbers into the initial numbers. We expect around 2/3rds to vote early, and that is exactly how it is. There is however a perception that 'Democrats vote early, independents and GOP vote on the day.' But hold on, we're still a week until election day. This indicates a huge turnout increase. We have had 10 days of early voting. 6 remain.
So we must discuss turnout trend. Thankfully NC have an excellent interactive that shows this. You can see them here:
https://www.ncsbe.gov/results-data/election-results/voter-turnout-statistics
Nobody has a crystal ball, but you'd expect this to translate to 1 million extra early voters statewide, minimum. Consider the total NC turnout was 4.7M in 2016, and we're going to have a minimum of 4.2M early voters in 2020, you can see Trump is quite fecked. (Unless these are the mystical shy trump voters). If we trust the polls (that 1/3rd will vote on election day), this would give us a total turnout of 6.3M, a 1.6 million voter gain on 2016. That's insane.
So lets break it down further. But first we'll slightly modify set 2 to discount New Hanover county. The reason for this is it's an outlier in the amount of support trump has there, relative to the other 5. So now our set 2 consists of 5 counties, which trump carried with between 57-67% of the vote in 2016. 340k have voted, with an average turnout of 44% (so far). 460k voted in 2016. 295k for trump. *165k for clinton
We'll make the assumption, that some of the extra 25% turnout will be in these set2 counties. Perhaps not as much as they are red, so we'll calculate a conservative 15% turnout increase in total.
That means, for these counties :530k will vote. Some 400k of these will vote early. (more based on our early voting data, less based on pollsters proportions) 40% are dems. 30% are republicans 30% are independents. These numbers are statewide and not county specific, but they would translate to a 280k early vote for Biden, 120k for Biden. More than trump got in total last time. It also means that the 1/3rd who vote on the day (lets be kind and say of the 180k instead of 130k vote on the day, Trump would need to carry almost every vote. You can clearly see, even if these numbers skew by 20%, Trump is pretty fecking screwed.
So in summary to this ramble:
- I think polls are misreading turnout and the young demographic especially.
- Biden is doing good in NC.
- Trumps attempts to mess with the USPS have backfired. (except for when I order stuff which then takes 2 months to arrive. It's definitely worked at messing that up)
[-Tillis is fecked]
- This post was gonna cover demographics too, but it became too long.